Closed iandees closed 7 years ago
The problem is that my file listing script in #92 was looking for files that ended in .tar.gz
. But it looks like some files (specifically the ones in Alaska?) end in .tar
without any compression.
The website lists 2,462 mosaic tiles, the tile index shapefile lists 9,036 tiles, and the lftp
file listing on the ftp site used in #92 listed just under 9,000. I don't really understand why all these numbers are different, but I'm going to use the file that lists the most data (the Shapefile tile index).
I opened the index Shapefile in Qgis, copied the attribute table to my text editor, and extracted the file URL out by itself. That file is in a gist here.
I'm transcoding the stuff we didn't get before (the stuff ending in .tar
) with:
cat pgc-updated-listing.txt | grep .tar$ | \
while read url; do
bn=$(basename ${url%.tar})
AWS_PROFILE=openterrain AWS_ACCOUNT_ID=273461098965 make submit-job job=aws/transcode-job.json.hbs input=${url} output=s3://elevation-sources-transcoded/pgdc_5m/${bn} name=$(sed 's/\./_/' <<< $bn)
done
The transcoding finished last night, but it looks like the new data that was added in Russia for release 6 ends with .tar.gz
so my transcode missed that data:
Purple is footprints database, green hashed polygons are footprints in the index shapefile for the dataset.
Release 6 is in the red here:
From ArcticDEM website.
Going to pull those in now.
I extracted the URLs for the data in release 6 mentioned above and put it in the above gist. Raw file here.
Using the same job submit script from above.
Finished transcoding, but also noticed that a swath of ~50 tiles are missing from their server:
I'm hoping to hear back from them some time this weekend.
The missing data was uploaded this morning and I transcoded it.
This data is all good to go in the database now, so I'm going to close this particular issue.
The ArcticDEM explorer website seems to show a whole bunch of empty spots in Alaska that weren't there before. Specifically, Denali seems to be missing lots of important detail:
...and the data downloaded in #92 seems to be missing all of Alaska:
The FTP URL we downloaded from in that ticket (
ftp://ftp.data.pgc.umn.edu/elev/dem/setsm/ArcticDEM/mosaic/v2.0/
) seems to have a ton of very recent last-updated timestamps, so I'm thinking that at the time I processed it they just didn't have everything posted?